Innovative analytical tools in the biopharmaceutical development : Applying SPR/Biacore, calorimetry and light scattering methods to the study of interactions between anticancer antibodies and the EGFR

Targeting of the epidermal growth factor receptor (EGFR) has become an established antitumor strategy with anti-EGFR antibodies approved for clinical use or in late stages of development. Postulation of antibody effector mechanisms has been based on in vivo or cell studies. These need to be complemented by an understanding of antibody/EGFR interactions on the molecular level. Thereon, crystal structures of the Fab fragments from different inhibitory antibodies in complex with the extracellular regions of EGFR have enlightened the molecular basis behind antibody-mediated EGFR inhibition. This study was focused on the further in vitro characterization of antibody/EGFR complexes in terms of stoichiometry, kinetics and thermodynamics of binding. Surface plasmon resonance (SPR)/Biacore, isothermal titration calorimetry (ITC) and static light scattering (SLS) were the tools employed to characterize the interactions between anticancer monoclonal antibodies and the epidermal growth factor receptor (EGFR). Clear stoichiometric evidence is provided for the binding of the monoclonal antibodies matuzumab, cetuximab and panitumumab to EGFR. These three antibodies are able to bind two EGFR molecules simultaneously, thus forming heterotrimer complexes. Independency of the two simultaneous EGFR binding events to one antibody molecule was confirmed with both kinetic and thermodynamic evidence. Unexpected stoichiometry results obtained for the nimotuzumab/EGFR interaction strongly indicate partial inactivity of the binding sites of this marketed antibody solution. Kinetically, the strong affinities of cetuximab and panitumumab could be related, respectively, to fast association and slow dissociation rates for the interactions of these two antibodies with EGFR. Similarly, the lower affinity of matuzumab could be assigned to a very fast dissociation of the matuzumab/EGFR complex. As for nimotuzumab, the lower affinity was mainly the result of a slower association rate to EGFR. Thermodynamically, the lower affinity known for matuzumab could be assigned to a higher entropic penalty upon binding. Interestingly, similar strong affinities of cetuximab and panitumumab were resolved to somewhat different thermodynamic profiles. Respectively, cetuximab interaction involves a higher enthalpy change compensated by an entropic penalty, while panitumumab interaction involves the lower enthalpy contribution of all four antibodies and an entropy change close to zero. All antibody/EGFR interactions were enthalpy-driven with either an entropy penalty or an entropy change close to zero. In contrast, interactions of the agonistic ligands EGF and TGF-alpha with EGFR were entropy driven and enthalpy penalised. Such different thermodynamic profiles are indicative of different binding processes for inhibitory antibodies and agonistic ligands. Motivated by reports on synergetic effects of the combined use of different EGFR-targeting antibodies, studies of the interdependent binding of antibody combinations to EGFR delivered insights into allosterism and relative epitope mapping. Results presented strongly corroborate the simultaneous binding of the antibody combinations matuzumab/cetuximab and matuzumab/nimotuzumab to EGFR. Concerning the binding of the combinations matuzumab/panitumumab, nimotuzumab/cetuximab and nimotuzumab/panitumumab to EGFR, displacement of the first antibody present upon binding of the second was observed. SPR results indicate that displacement of the first antibody could be caused by small conformational shifts upon binding of the second antibody. The applicability of the biophysical methods used for the generation of meaningful quantitative data on binding interactions is demonstrated. Furthermore, a comparative assessment of the biophysical tools SPR, ITC and SLS to the study of protein-protein interactions is presented. The possibility of real time monitoring of the interactions was a special feature of SPR that enabled determination of the binding kinetics. Since SPR analysis involves immobilization of one interactant, it is not necessarily representative of what happens in solution. However, thermodynamic characterization of antibody interactions with EGFR performed with SPR delivered enthalpy and entropy changes that correlate well with ITC results. In fact, EGFR being a membrane protein, the adequacy of a solution method such as ITC to be more representative of the in vivo situation than a surface method such as SPR could be contested. The results presented rather demonstrate the combined utilities and corroborative use of SPR and ITC, with SLS providing an additional qualitative confirmation of the assembly states